- Minor bugs fixes - UX improvements

- Minor bugs fixes - UX improvements

- Minor bugs fixes - UX improvements

- Minor bugs fixes - UX improvements

- Minor bugs fixes - UX improvements

- Improved AI harness for better understanding and more accurate execution of user requests.

-Prevented messages from being sent before attached images finish uploading. -Added protection against duplicate submissions. -Fixed conversation reload issues that could reuse old Q&A IDs. -Improved conversation loading by merging data correctly and removing duplicate legacy messages. -Failed image generation jobs are now saved with detailed error information. -Added automatic retries for temporary connection and image download failures. -Removed misleading generic error messages when image generation continues successfully in the background. -Added localized upload status messages. -General stability improvements and bug fixes.

- Minor fixes - Better user experience - 8 new PDF Agents

- Minor fixes - Better user experience - 8 new PDF Agents

- Minor fixes - Better user experience

- Minor fixes - Better user experience

- Improved UX - Minor bug fixes

- Voice agent bug is resolved - Minor bug fixes

- Smarter AI with improved reasoning and execution. - Enhanced AI experience for faster, more reliable results.

- Added a new prompt field to AI agents. You can now provide simple instructions before running an agent, helping the AI deliver more accurate and tailored results. - Improved UI compatibility across different iPhone models. Fixed layout issues on smaller devices, including iPhone SE, for a more consistent experience.

- AI-powered product management is here. Folio AI can now create and update products, manage features, organize lists, attach media, reorder product images, create stores, and more. - Folio AI can now reason through multiple steps and complete complex workflows in a single request. - Added HTML generation. Generated content can be previewed in the app and exported as PDF. - Redesigned the AI Assistant home screen with a cleaner experience and quick access to PDF export. - Simplified AI tool selection by grouping related image and video generation tools into a more intuitive interface. - Added support for the latest video generation models within the AI Assistant. - Improved agentic workflows and context management for more reliable multi-step AI execution. - Added new UI experiences for displaying tool results. - Introduced a new reasoning and planning engine. Folio AI can now analyze tasks, plan execution, and intelligently complete workflows when needed. - Removed the legacy deterministic reasoning engine in favor of the new AI-powered planning architecture. - Store slugs can now be updated directly from Account Settings. - Users with generic organization slugs are now prompted to choose a custom store URL before publishing a store.

- Turn any product list into a beautifully designed PDF catalog in seconds — ready to share with customers, buyers, and partners.

- Improved AI credit display. Users can now see when credits will renew, along with a clear breakdown between purchased credits and plan-based credits. - Added an upgrade path in Account Settings, making it easier to move to a higher subscription plan. - Added upgrade options within the Additional Credits purchase flow, allowing users to upgrade their plan instead of purchasing extra credits. - Improved Folio AI knowledge retrieval. When users ask questions about Folio, the assistant now automatically uses dedicated Folio knowledge sources to provide more accurate and consistent answers.

- Agent templates now support multi-select items with an improved selection experience. - Agent parameters are now displayed in user messages and stored in conversation metadata for better transparency. - New Fashion Model Agent: collects age, gender, and brand positioning information, then generates four distinct fashion model alternatives. - New Virtual Dressing Agent: creates professional full-body fashion photos by dressing a reference model with uploaded clothing items. - New Model Animation Agent: transforms studio fashion model photos into premium fashion videos while preserving the model's identity and garment details. - Improved image selection. Both agent templates and the main AI agent can now use images generated in previous AI jobs. - Fixed an issue where the sound on/off parameter was not working correctly in video agents. - Video thumbnails are now cached for improved performance and faster loading.

- Smarter image follow-ups: Folio now understands previously generated images as editable assets, can reason about which image you mean, and asks for clarification when multiple source images are available. - Improved agent context during execution, helping the AI better understand user intent. - Order management is now available in the mobile app: + View orders + Update order status + Reply to customer messages + Edit products within an order - Improved subscription (paywall) design. - Added a new monthly subscription plan. - Agent runs now appear in conversation history using the localized agent name instead of a generic system message. - Enhanced analytics tracking for agent usage, tools, and memory features. - Refactored tool-calling architecture with a standardized structure that separates tool parameters from metadata. - Added summaries for image generation and image editing results, now displayed as image captions. - Added captions for video models. Generated videos now display a short summary beneath the video. - The 4K option has been integrated into the unified tool parameter architecture. - When users have more than three product lists, all lists are now grouped under a Product Lists section for a cleaner homepage. - Added a 4K option directly to tool controls in Folio AI. - Improved tool controls with clearer parameter labels. - Better visual guidance for tools that require image uploads. - Style memory controls now include visual cues for image uploads. - Tool controls and style memory controls now follow a unified design language. - Added an approval workflow to conversations. Before image or video jobs run, users can review tool calls and parameters, then approve or cancel each job. - For multi-asset generation agents, users can choose exactly which assets should be created. - Pending jobs now appear in conversations with a waiting-for-approval status. - Smarter AI awareness of cancelled jobs. Cancelled actions are stored in conversations, displayed in the UI, and included in future context. - Added automatic approval preferences for future jobs, configurable from Account Settings.

- Smarter image follow-ups: Folio now understands previously generated images as editable assets, can reason about which image you mean, and asks for clarification when multiple source images are available. - Improved agent context during execution, helping the AI better understand user intent. - Order management is now available in the mobile app: + View orders + Update order status + Reply to customer messages + Edit products within an order - Improved subscription (paywall) design. - Added a new monthly subscription plan. - Agent runs now appear in conversation history using the localized agent name instead of a generic system message. - Enhanced analytics tracking for agent usage, tools, and memory features. - Refactored tool-calling architecture with a standardized structure that separates tool parameters from metadata. - Added summaries for image generation and image editing results, now displayed as image captions. - Added captions for video models. Generated videos now display a short summary beneath the video. - The 4K option has been integrated into the unified tool parameter architecture. - When users have more than three product lists, all lists are now grouped under a Product Lists section for a cleaner homepage. - Added a 4K option directly to tool controls in Folio AI. - Improved tool controls with clearer parameter labels. - Better visual guidance for tools that require image uploads. - Style memory controls now include visual cues for image uploads. - Tool controls and style memory controls now follow a unified design language. - Added an approval workflow to conversations. Before image or video jobs run, users can review tool calls and parameters, then approve or cancel each job. - For multi-asset generation agents, users can choose exactly which assets should be created. - Pending jobs now appear in conversations with a waiting-for-approval status. - Smarter AI awareness of cancelled jobs. Cancelled actions are stored in conversations, displayed in the UI, and included in future context. - Added automatic approval preferences for future jobs, configurable from Account Settings.

- Added a new Feedback section in Account View. User feedback is now sent directly to our support email. - Memory images are now stored in smaller sizes for better performance and storage optimization. - The new Store system is now live on iOS: +Create and edit stores +Enable ordering +Select price lists +Update and reorder lists +Add or remove lists from existing stores - Catalogs published directly from lists will gradually be deprecated. Old catalog links are removed from the main page but still accessible from Account View. - Price list cards added to the My Folio page, making it easier to create and manage pricing. - Product prices can now be edited directly from product detail pages. - Added loading states after agent runs to prevent duplicate actions, especially on slow connections. - Fixed stop button behavior in conversations. - Fixed an edge-case issue with conversation ID assignment by improving serialized conversation allocation per user and agent.

- Swatch Agent updated: create consistent studio-shot product images from any uploaded swatch. Users can now select previously-created studio style memories and apply the same visual style across their entire catalog. - New Studio Style Memory Agent: analyze a product image and automatically create reusable studio styles including background, lighting, shadows, depth, focus, product angle, finish state, and aspect ratio. - Memory section added to the main AI agent. Users can now quickly access, apply, manage, edit, or delete previously-created memory styles directly from conversations. - Memory structure improved with support for stored reference images. - Device-ID users who have not added an email now see an update email prompt with web app access information. - Folio AI information context improved: the assistant now uses a dedicated markdown knowledge file for better responses about Folio. - Swatch Agent improved with a new option to use uploaded images directly as the base image without converting them into studio shots. - New Reference Style Memory Agent: create reusable memories from reference images or directly from generated/edited conversation images. - Analytics improved across Folio AI: better tracking for tool usage, memory creation, and memory reuse patterns. - Catalog terminology cleanup completed and duplicate catalog functions removed. - iOS API performance optimized: reduced duplicate endpoint calls with smarter request handling and caching. - Catalog APIs fully refactored and cleaned up — now significantly smaller and more maintainable. - Device-ID login system strengthened with improved recovery flow, fewer unexpected sign-outs, and secure Keychain-based device storage. - Authentication performance improved with optimized onboarding, purchase, and profile flows. - Attachment button redesigned: image icon replaced with a plus button that opens image upload, camera, and product selection options. - Bug fixes and stability improvements: - Published catalogs now correctly appear in My Folio even for users with fewer than 3 products. - Existing account/device login issue creating duplicate accounts has been fixed. - Duplicate conversation generation issue resolved. - Main AI agent display issue on iPad fixed.

- Tüm ajanlar yeni bir mimariye taşındı. - Kimlik doğrulama hataları için görüntü düzenleme akışı, daha kullanıcı dostu hata mesajları ve ek yeniden deneme mantığı ile iyileştirildi. - Üstel geri çekilme kullanarak geçici 401, 408, 429, 5xx ve ağ hatalarını ele almak için yeniden kullanılabilir bir API yeniden deneme yardımcısı tanıtıldı.

-If the sample image is not changed in agent input, users are now asked to confirm before continuing with the sample image. -For flexible image inputs, additional images are dimmed and inactive until the first image is replaced, with a guiding hint shown to the user. -The history drawer sorting issue is fixed; instead of prioritizing unviewed jobs, items are now sorted purely by history. -Agents in the Essentials category are moved to the main agent area, and their parameters now appear inline when selected. -Recently used agents are moved to the main agent area for quicker access. -Users with invalid email addresses are now prompted to update their email directly in the main agent area below the greeting section. -The AI Assistant page is now set as the new landing page.